基于贝叶斯网的高维数据隐藏模式挖掘

High Dimensional Data Hiding Pattern Mining Based on Bayesian Network

针对当前高维数据隐藏模式挖掘精度较低、执行时间较长,且挖掘过程工作量较大,过程较为复杂的问题,提出了基于贝叶斯网的高维数据隐藏模式挖掘方法。通过有向无环图像与概率表构成贝叶斯网络,分析数据挖掘框架,采用贝叶斯网络对高维数据缩小开销计算,利用信号处理方法提取数据信息特征,对高维数据信息子空间降维,采用自适应级联滤波完成数据的降噪,将多通道的传感信息数据构成自适应的波束并完成聚焦,以此完成高维数据的隐藏挖掘。实验结果表明,所提方法的高维数据隐藏模式挖掘执行时间短,能够有效提高数据挖掘精度,且挖掘过程工作量较小,挖掘过程较为简单。

Currently,there are some defects in high-dimensional data hiding pattern mining method,such as low precision,long execution time,heavy workload,and complex process etc..In this regard,a mining method of high-dimensional data hiding patterns based on Bayesian network is proposed in this work.Firstly,Bayesian network was constructed by directed acyclic image and probability table to analyze the framework of data mining for reducing the cost of high-dimensional data.Then,the feature of data information was extracted via using signal processing method to reduce dimension and high dimension data information subspace.Finally,the adaptive cascade filter was utilized to denoise the data,and an adaptive beam was generated after using the multi-channel sensing information data and the focusing was achieved,thus the hidden mining of high-dimensional data was completed.The simulation results show that the method has the advantages of short mining time,high precision,small workload and simple process.